Dirac live vs Acourate vs Audiolense

[edorr]

Do yourself a favor: Stop with the DLNA!! It's a joke standard.

AES67 IS THE STANDARD ethernet protocol. You should be looking for devices that support AES67, like Ravenna. You can do everything you want to do with a Merging Hapi over ethernet.

I presume the new PS audio bridge would not support it, so to get best sound quality I would need an AES67 to I2S (PS audio interface) converter. The Sonore Signature Series Rendu does not support that either. Seems like this protocal may be a bit too far ahead of the curve for practical implementation for some time.

I'll look into the hardware / software you are mentioning though. I have not been keeping up to date. Not quite ready to change DAc hardware though.

 

[Brucemck2]

I have been able to make J River apply an Acourate convolution file when streaming via dlna

I have not been able to get it to do both sample rate conversion and convolution.

In my case, what I do is convert the files offline (both the PCM conversion and application of the Acourate filter) using AcourateNAS, so that it's very simple to do dlna streaming. (Basically, I've created a "duplicate" library on my NAS, which is not a big deal given the low cost of hard drive storage.) Alternatively, you can PCM convert only those files that will require it, and use Acourate convolution files while streaming.

 

[dallasjustice]

Keep it simple

I presume the new PS audio bridge would not support it, so to get best sound quality I would need an AES67 to I2S (PS audio interface) converter. The Sonore Signature Series Rendu does not support that either. Seems like this protocal may be a bit too far ahead of the curve for practical implementation for some time.

I'll look into the hardware / software you are mentioning though. I have not been keeping up to date. Not quite ready to change DAc hardware though.

The PSA gear take AES. If you are stuck on i2s then maybe that's the only solution. I'm just sayin' you wouldn't need to have 2 totally different systems. I know a dude on Acourate forum with a Merging Hapi he feeds to an MSB DAC and get surround sound with all of the other channels out of the Hapi. That's the right way to do this. That way, you are future proof on the ethernet and you can sync up all channels in/out so that you can take perfect measurements in Acourate or Audiolense. It would be MUCH simpler.

I am prolly going to put my money where my mouth is on the Hapi. I will use it as my DAC and ADC. It can do up DSD256 so it should be fun to mess around a little with HQPlayer using some filters I created in Audiolense.

Other than the ability to do tons of MCH, Ravenna ethernet uses an ASIO protocol driver which can tightly synchronize multiple DAC/ADCs on the same network switch. That may not seem like it's important now, but I bet it will be important in the near future. Here's a video of some Genelec audio over IP speakers demo.

 

[edorr]

The PSA gear take AES. If you are stuck on i2s then maybe that's the only solution. I'm just sayin' you wouldn't need to have 2 totally different systems. I know a dude on Acourate forum with a Merging Hapi he feeds to an MSB DAC and get surround sound with all of the other channels out of the Hapi. That's the right way to do this. That way, you are future proof on the ethernet and you can sync up all channels in/out so that you can take perfect measurements in Acourate or Audiolense. It would be MUCH simpler.

I am prolly going to put my money where my mouth is on the Hapi. I will use it as my DAC and ADC. It can do up DSD256 so it should be fun to mess around a little with HQPlayer using some filters I created in Audiolense.

Other than the ability to do tons of MCH, Ravenna ethernet uses an ASIO protocol driver which can synchronize multiple DAC/ADCs. That may not seem like it's important now, but I bet it will be important in the near future. Here's a video of some Genelec audio over IP speakers demo.

Seems like I am way behind the curve and need to do some catching up. This could get complicated. I'll also ask Ted Smith.

When you say "PSA gear take AES", does this mean you can stream AES protocol to a PS audio bridge over ethernet and it will work? If I have three PS audio DACs for MCH, would the protocol clock sync between these DACs?

 

[Vincent Kars]

However, I need to keep DSP functionality. I am currently using CAPS 3.0 -> JRiver -> Dirac ->USB. Can anyone confirm if I could redeploy the server as a DLNA streamer, running Acourate convolusion instead of Dirac? In addition to PEQ, the Jriver DSP should also do DSD to PCM conversion for my 2 channel SACD rips.

Am I correct in assuming this (DLNA streaming / acourate) should sound better because of (A) superior DSP from Acourate (B) inherent superiority to I2S bridge over USB input on the Directstream DAC?

As far as I know, in JRiver 20 you can configure the DLNA server. You have all the options available in DSP studio at your disposal.

Recently I had a quick look at Ravenna http://thewelltemperedcomputer.com/HW/Connect/Prosumer.htm
The concept looks very nice indeed.
You have a virtual soundcard at the PC side so you can do all the DSP you want and send it to this virtual card.
Looks very flexible to me.
If I remember correctly, Bruce B is using this protocol.

I wonder if I2S is superior to async USB.
The question is who does the clocking and the accuracy of this clocking.
If in both cases the clocking is done by the DAC I wonder if there is any difference between this protocols.

 

[edorr]

As far as I know, in JRiver 20 you can configure the DLNA server. You have all the options available in DSP studio at your disposal.

Recently I had a quick look at Ravenna http://thewelltemperedcomputer.com/HW/Connect/Prosumer.htm
The concept looks very nice indeed.
You have a virtual soundcard at the PC side so you can do all the DSP you want and send it to this virtual card.
Looks very flexible to me.
If I remember correctly, Bruce B is using this protocol.

I wonder if I2S is superior to async USB.
The question is who does the clocking and the accuracy of this clocking.
If in both cases the clocking is done by the DAC I wonder if there is any difference between this protocols.

May be I should wait until the verdict is in on USB vs. I2S bridge interface. I was intrigued by the DirectStream review on computeraudiophile, that said using the Sonore Signature Series Rendu for Ethernet to I2S conversion was by far the best path to get best sound out of the Directstream. The bridge does essentially the same as the Sonore Signature Series Rendu - convert ethernet to I2S, so I figured it would be superior to USB...

 

[Brucemck2]

As far as I know, in JRiver 20 you can configure the DLNA server. You have all the options available in DSP studio at your disposal.

The interface options suggest that is the case, but in practice it does not work (yet):
http://yabb.jriver.com/interact/index.php?topic=92745.msg642933#msg642933

 

[dallasjustice]

The problems with DLNA:
1. no highend audio hardware which can do MCH
2. limited DSP/convolution functionality in Jriver
3. No ASIO complaint drivers; therefore, no possibility to use DLNA for measurement.

Since this thread is mostly about DSP software like Acourate, Audiolense and DIRAC, these limitations are such that one should ask the question: Why use DLNA when there's a much better ethernet standard available without any of these limitations? This is an open protocol; there's nothing proprietary about it.

Erik,
The Hapi has AES/EBU outs as well as a variety of other digital outs. So, you could easily send your music over ethernet to the Hapi and then go out from the Hapi to AES/EBU into your DAC, if you choose to do that.

I'll be able to give more feedback later on concerning the quality of the Hapi's DtoA, but I'm told its one of the best DACs. We will see.

 

[dallasjustice]

I had an awesome Teamviewer session with Uli of Acourate today. For those who don't know Uli, you should learn more about him and Acourate.

I used OBS to record the Teamviewer session because Uli demonstrated several advanced Acourate features which are often the subject of extensive discussion and investigation. I asked Uli to demonstrate his method for incorporating subs with RL main speakers. It turned out to be a marathon session (good thing I sent Uli some money beforehand to cover some of his time) because he wanted to show me some things I didn't expect going into it. I will post the OBS monitor capture videos later on my Vimeo account. But I'll need to edit them into four parts. The lessons I learned today:
1. Why dual mono subs are almost impossible to time align.
2. Why Toole's recommendation to place two mono subs one quarter width on front wall isn't appropriate in my case.
3. Why simply using an impulse isn't sufficient to time align and can be the worst scenario for sub integration. Sub integration can't be automated.
4. How to use corner frequency sine waves to better time/phase align subs to main speakers.

Number 4 has been a topic of discussion recently and I believe the OBS video will greatly help others understand Uli's method. I believe his method is the best possible way to integrate subs and the results speak for themselves.

Acourate is back in my life. I really missed the beautifully articulate mid/hi and HF range. And, of course, I've got that BASS! As much as I love Auidiolense, Acourate clearly reaches into another dimension.

 

[Brucemck2]

That's one session I'm looking forward to seeing! Integrating subs seamlessly isn't easy.

I don't use Trinnov's automated routine, but, have found the automated suggestions for distance and phase to provide helpful starting points. They don't disclose how they do it, but their distance compensation for subs often doesn't match their measured distance. I was told their algorithm looked at phase alignment through a likely crossover region.

 

[Brucemck2]

I had an awesome Teamviewer session with Uli of Acourate today. For those who don't know Uli, you should learn more about him and Acourate.

1. Why dual mono subs are almost impossible to time align.

Can you expand on this? Would it be true for four subs placed symmetrically with regard to the front L/R mains and room boundaries, or, is it meant to capture asymmetries from multiple subs, or something else?

 

[Fitzcaraldo215]

That's one session I'm looking forward to seeing! Integrating subs seamlessly isn't easy.

I don't use Trinnov's automated routine, but, have found the automated suggestions for distance and phase to provide helpful starting points. They don't disclose how they do it, but their distance compensation for subs often doesn't match their measured distance. I was told their algorithm looked at phase alignment through a likely crossover region.

Measured linear distance not equal to mike determined distance is quite common for subs. If the subs have a built-in DSP or other form of delay in their input control setups, such as a phase control, then, the sub may be adding delay to the signal. It seems the majority do these days. Often this cannot be eliminated. So, the mike measurement is likely a more accurate determination of the true acoustic delay = effective distance. It can possibly be tweaked further for proper phase with mike measurements using the Seton method: parametrically varying phase or the controller's distance setting to achieve maximum measured output in the xover region.

 

[dallasjustice]

Can you expand on this? Would it be true for four subs placed symmetrically with regard to the front L/R mains and room boundaries, or, is it meant to capture asymmetries from multiple subs, or something else?

I should have been more specific. My subs are both equidistant to seated position on the frontwall about 7-8 feet behind my main RL speakers.

It's not possible to time and phase align 2 mono subs with main RL speakers, using the more precise method Uli uses in my room. Uli's delay calculation uses both HF impulse from RL tweeter and low frequency impulse from sub. But he takes things several steps further than any automated system I know of. He generates a perfect sine wave at the corner frequency and then lines it up with the subwoofer at that frequency. He then makes an adjustment to account for best possible phase alignment between the perfect 80hz tone and the sub. This is where things can get a little messy with mono subs. The mono subs just don't look as clean in the phase to be able to make a coherent adjustment. And when we were able to make such an adjustment to account for the phase, the delays we came up for the two subs were off by quite a bit when they were in mono even though they are exactly the same distance from the seat. We did NOT experience this problem when he setup the stereo sub configuration. The delays were off by only 20 samples (48khz logsweep) or so; very close.

The final step in the process after the RL and subs are time/phase aligned, is to take a final stereo measurement using the crossovers convolved with the phase/time alignment settings. This way, the final correction is applied to the ACTUAL combined in room response. I would consider this process to be very manual and requires some skills. Obviously, I am just learning myself, but I know enough now so that I feel I can give an opinion as to why I think Uli's method is superior to any other method I've seen to integrate subs. The videos will go into much great depth than what I'm describing here and I think they could be helpful to others. I am not good with video editing, so it may take me a few days to get everything posted.

Michael.

 

[dallasjustice]

Measured linear distance not equal to mike determined distance is quite common for subs. If the subs have a built-in DSP or other form of delay in their input control setups, such as a phase control, then, the sub may be adding delay to the signal. It seems the majority do these days. Often this cannot be eliminated. So, the mike measurement is likely a more accurate determination of the true acoustic delay = effective distance. It can possibly be tweaked further for proper phase with mike measurements using the Seton method: parametrically varying phase or the controller's distance setting to achieve maximum measured output in the xover region.

That correct.

But things can get pretty complex when you start measuring. One of the issues we ran into in my case is that my speakers are three way passive. Although the tweeter/mids are connected in positive polarity, the woofer was intentionally designed to be connected in negative polarity. So, distance is NOT a very good method for calculating best possible delay in my case. I know alot of other speakers are designed this way so it's impossible to come up with a universal answer. It turned out that the delays we used were almost one half of what one would have predicted if one were to just look at the distance. Once we understood the woofer was connected in negative polarity, it all made sense to me why we were getting much less delay than I predicted we should need.

 

[Fitzcaraldo215]

That correct.

But things can get pretty complex when you start measuring...

I agree completely. That is why I don't do it. Yes, I believe that measurements are absolutely crucial, and I trust my DSP EQ tool, Dirac in my case, to do them in a sophisticated way consistent with their design, if not perfectly. But, I have seen many a thread where the availability of good independent measurement tools has led countless afficianados down the slippery road to who knows where. A little bit of knowledge is a dangerous thing is my credo. That is me about room acoustics, measurements, etc. Given a choice between just relaxing, putting my feet up and listening to about the best music reproduction I have ever heard anywhere compared to live vs. going down the rabbit hole and digging ever deeper to tweak it further, why, it is no contest for the pragmatist in me.

I am not taking a swipe at you. You have shown in countless postings here and elsewhere that you are willing to do whatever it takes to learn completely about what is going on and to achieve a level of understanding and perfection that others can only dream of. And, you are just as caring about sharing this knowledge with others, here and elsewhere. I respect and admire your efforts.

Maybe I am just lazier than you. But, that sound that I have got now just seduces me to want to listen and not fiddle.

Best to you,

Carl

 

[dallasjustice]

I agree completely. That is why I don't do it. Yes, I believe that measurements are absolutely crucial, and I trust my DSP EQ tool, Dirac in my case, to do them in a sophisticated way consistent with their design, if not perfectly. But, I have seen many a thread where the availability of good independent measurement tools has led countless afficianados down the slippery road to who knows where. A little bit of knowledge is a dangerous thing is my credo. That is me about room acoustics, measurements, etc. Given a choice between just relaxing, putting my feet up and listening to about the best music reproduction I have ever heard anywhere compared to live vs. going down the rabbit hole and digging ever deeper to tweak it further, why, it is no contest for the pragmatist in me.

I am not taking a swipe at you. You have shown in countless postings here and elsewhere that you are willing to do whatever it takes to learn completely about what is going on and to achieve a level of understanding and perfection that others can only dream of. And, you are just as caring about sharing this knowledge with others, here and elsewhere. I respect and admire your efforts.

Maybe I am just lazier than you. But, that sound that I have got now just seduces me to want to listen and not fiddle.

Best to you,

Carl

I totally understand where you are coming from. Thanks for the kind words here and elsewhere.

 

[Brucemck2]

The final step in the process after the RL and subs are time/phase aligned, is to take a final stereo measurement using the crossovers convolved with the phase/time alignment settings. This way, the final correction is applied to the ACTUAL combined in room response. I would consider this process to be very manual and requires some skills.

So, you get distance/time delays and phase by using a sine wave at the corner frequency. So far so good.

Then I got confused ...

You take a Left measurement and generate a Left correction file, where during measurement you run the left main and left sub using the delays and crossover generated earlier, and then repeat that for the right channel, or

You generate a Left main speaker measurement and correction curve, a right main speaker measurement and correction curve, and then run the subs together and generate a single subwoofer correction curve that gets applied uniformly to the two subs. After that, during playback, you apply the low pass and high pass crossovers

Or something else?

As an aside, I admire your tenacity in getting to technically superior solutions.

 

[dallasjustice]

Acourate Subwoofer Setup Video

This is an example from my room. This is a TeamViewer session a few days ago with Uli in control of my PC. This video demonstrates a few different acourate functions. I've edited this down and there was an hour of work we did with mono subs before this. I will post that one later because there are many good lessons in that one too. There are a few things at the beginning you may not be interested to see. Most importantly, Uli shows how he setups subs in my 4CH, stereo sub system.

[video=vimeo;124667294]https://vimeo.com/124667294[/video]

https://vimeo.com/124667294

 

[Brucemck2]

The final step in the process after the RL and subs are time/phase aligned, is to take a final stereo measurement using the crossovers convolved with the phase/time alignment settings. This way, the final correction is applied to the ACTUAL combined in room response.

The Vimeo was very helpful. It's a great approach for integrating stereo subs into a stereo system.

Have you measured the response (both speakers running at the same time) at the low end post correction/convolution? (That is a proxy for low frequency close-to- mono information mixed uniformly in the left and right channels, like a pipe organ often is.) In my room running stereo bass the left and right measure individually great into the mid 30s, but when combined as a mono signal I get some phase issues that causes some roll off from the lower 50s. I don't hear any issues, and my bass sounds full and impactful, but it measures less than ideal. I'm considering trying to address this by running the mains (fully convolved) full range, and then configuring a mono sub channel who's only job is to fill in that very low end by a few db. It remains to be seen if I can do that without screwing up the individual left and right for non-mono less uniformly shared low end signals (like bass guitar harmonics). If I can't, then I may create some automation where I could switch those extra subs in solely for a few select mono low end heavy recordings.

 

[dallasjustice]

I haven't measured yet. The method I need to use to measure goes through Jriver ASIO line in. I would need to make a cfg file and prolly rename the .wav impulses. I'm not looking forward to that.

The Vimeo was very helpful. It's a great approach for integrating stereo subs into a stereo system.

Have you measured the response (both speakers running at the same time) at the low end post correction/convolution? (That is a proxy for low frequency close-to- mono information mixed uniformly in the left and right channels, like a pipe organ often is.) In my room running stereo bass the left and right measure individually great into the mid 30s, but when combined as a mono signal I get some phase issues that causes some roll off from the lower 50s. I don't hear any issues, and my bass sounds full and impactful, but it measures less than ideal. I'm considering trying to address this by running the mains (fully convolved) full range, and then configuring a mono sub channel who's only job is to fill in that very low end by a few db. It remains to be seen if I can do that without screwing up the individual left and right for non-mono less uniformly shared low end signals (like bass guitar harmonics). If I can't, then I may create some automation where I could switch those extra subs in solely for a few select mono low end heavy recordings.